ORLANDO — Hopes that the addition of bevacizumab to gemcitabine would improve survival in patients with advanced pancreatic cancer have been dashed by the disappointing results of CALGB 80303, a phase III randomized trial presented at a symposium on gastrointestinal cancers sponsored by the American Society of Clinical Oncology.
An earlier phase II trial of gemcitabine—the standard drug for advanced pancreatic cancer—plus bevacizumab in 52 patients showed responses in 21% and a median survival of 8.8 months. These results were far better than the 5%–10% response rate and 5–6 month survival expected with gemcitabine alone, and prompted the current trial, said lead author, Dr. Hedy Lee Kindler, director of gastrointestinal oncology at the University of Chicago. Unfortunately, the study did not confirm the promising data of its predecessor, she said.
The CALGB (Cancer and Leukemia Group B) 80303 trial randomized 602 patients with advanced pancreatic cancer to gemcitabine plus bevacizumab or gemcitabine plus placebo, and followed them for overall survival, objective response rate, progression-free survival, and toxicity.
The patients were enrolled from June 2004 to April 2006, and weeks later, the CALGB Data Safety Monitoring Board determined that a significant difference in survival between the treatment arms would be unlikely. As a result, all patients on treatment were notified and unblinded, and those who were thought to be benefiting from bevacizumab were allowed to continue the drug with informed consent, Dr. Kindler said at the symposium also sponsored by the AGA Institute, the American Society for Therapeutic Radiology and Oncology, and the Society for Surgical Oncology.
There were no statistically significant differences in survival (5.7 months for gemcitabine plus bevacizumab vs. 6.0 months for gemcitabine alone) or progression-free survival (4.8 months for gemcitabine plus bevacizumab vs. 4.3 months for gemcitabine alone). Treatment with gemcitabine plus bevacizumab resulted in a very slight increase in the rate of tumor shrinkage and stable disease, compared with gemcitabine plus placebo (54% vs. 47%, respectively), Dr. Kindler reported.
However, bevacizumab was associated with grade 3 and 4 hypertension and proteinuria.
The encouraging results in the earlier study may have occurred because that smaller study had proportionally more patients with good prognostic factors, Dr. Kindler said. She added that ongoing companion studies of CALGB 80303 may provide key insights into the biology of pancreatic cancer.
Dr. Charles A. Staley, chief of surgical oncology at Emory University, Atlanta, commented that the treatment of pancreatic cancer continues to be frustrating. “It's a little disappointing—bevacizumab has had a great run in most diseases—but again, this just points out the complexity of drug interactions and what actually happens in the host environment.”